app.py

import streamlit as st
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import os

# Try importing openpyxl, and install if not present
try:
    import openpyxl
except ImportError:
    st.error("openpyxl module is required to read Excel files. Attempting to install...")
    os.system('pip install openpyxl')
    st.success("openpyxl installed successfully. Please restart the app.")

# Import the Groq SDK (after installing it)
try:
    from groq import Groq
except ImportError:
    st.error("The 'groq' library is required but not installed. Please ensure it's included in requirements.txt.")
    raise

# Configure page
st.set_page_config(page_title="Data Augmentation App", layout="wide")
st.markdown(
    f"""
    <style>
        .reportview-container {{
            background: url("https://cdn.pixabay.com/photo/2016/06/02/02/33/triangles-1430105_1280.png");
            background-size: cover;
        }}
        footer {{
            text-align: left;
        }}
    </style>
    """,
    unsafe_allow_html=True,
)

st.title("Data Augmentation and Analysis App")
st.sidebar.title("Upload Your File")
st.sidebar.markdown("Supported formats: CSV, Excel")

# Fetch Groq API key from Streamlit secrets
groq_api_key = st.secrets.get("HUGGINGFACE_KEY")  # Fetch Groq API key stored in secrets.json
if not groq_api_key:
    st.error("Groq API key not found in secrets.")
else:
    st.success("Groq API Key Loaded Successfully!")

# Initialize Groq client
client = Groq(api_key=groq_api_key)

# Function to load data from file (CSV/Excel)
def load_file(uploaded_file):
    """Load the uploaded file."""
    if uploaded_file.name.endswith('.csv'):
        return pd.read_csv(uploaded_file)
    elif uploaded_file.name.endswith('.xlsx'):
        return pd.read_excel(uploaded_file)
    else:
        st.error("Unsupported file format. Please upload a CSV or Excel file.")
        return None

# Function to generate graph based on user query
def generate_graph(data, query):
    """Generate a graph based on user query."""
    try:
        fig, ax = plt.subplots(figsize=(10, 6))
        
        if "correlation" in query.lower():
            # Correlation matrix
            sns.heatmap(data.corr(), annot=True, cmap="coolwarm", ax=ax)
            st.pyplot(fig)
        elif "histogram" in query.lower():
            # Histogram
            column = st.selectbox("Select a column for the histogram", data.columns)
            sns.histplot(data[column], kde=True, ax=ax)
            st.pyplot(fig)
        elif "bar" in query.lower() and "country" in query.lower():
            # Bar chart for countries and units sold
            if 'country' in data.columns and 'units sold' in data.columns:
                country_data = data[['country', 'units sold']].groupby('country').sum().reset_index()
                sns.barplot(x='country', y='units sold', data=country_data, ax=ax)
                ax.set_xticklabels(ax.get_xticklabels(), rotation=45, ha='right')
                st.pyplot(fig)
            else:
                st.error("The dataset must contain 'country' and 'units sold' columns.")
        else:
            st.error("Unsupported graph type. Try asking for a correlation matrix, histogram, or bar chart.")
    except Exception as e:
        st.error(f"Error generating graph: {e}")

# Function to handle user queries using the Groq model
def handle_query(query):
    """Handle user query using the Groq model."""
    try:
        # Send user query to Groq model using Groq SDK
        chat_completion = client.chat.completions.create(
            messages=[{"role": "user", "content": query}],
            model="llama3-8b-8192",  # Model name as per your Groq model
        )
        
        # Display the response
        st.write("Response from Groq Model:", chat_completion.choices[0].message.content)
        
    except Exception as e:
        st.error(f"Error in Groq processing: {e}")

# Main App
uploaded_file = st.sidebar.file_uploader("Upload your file here", type=["csv", "xlsx"])
if uploaded_file:
    data = load_file(uploaded_file)
    if data is not None:
        st.write("Dataset Preview")
        st.dataframe(data)

        query = st.text_area("Ask your question about the dataset")
        if query:
            if "table" in query.lower():
                st.write("Table Preview")
                st.write(data)
            elif "graph" in query.lower():
                generate_graph(data, query)
            else:
                handle_query(query)  # Use Groq model for query processing

# Footer
footer = """
    <div style='text-align: left; padding: 10px;'>
        <footer>Created by: Shamil Shahbaz</footer>
    </div>
"""
st.markdown(footer, unsafe_allow_html=True)